HomeAtlasDashboardChartsMetalsResearchPersonsBubble MapGlobeLeadershipLensesMethodologyGlossarySource ↗
Independent research & opinion. Gradings are automated / LLM-assisted and may contain errors or hallucinations; nothing here is a statement of fact, financial advice, or an accusation of wrongdoing by any party. Claims about identifiable people or organizations reflect public records + good-faith interpretation; intent is not inferred from association. Methodology & disclaimer.

Critical minerals / rare earths — the physical bottleneck under AI + defense, and the state-capitalism circularity

Built from research/macro-critical-minerals.json (web-verified: MP Materials/Apple/DoD press + SEC 8-K, CNBC, Critical Metals/Tanbreez, TMC, China export-control reporting — Pillsbury, FDD, Fortune, CFR). Models: models/graph/defense_web.py, models/z3/defense_chokepoint.py. Equities cohort charted in the Charts tab.

Thesis. Rare earths / critical minerals are the physical chokepoint beneath BOTH the AI/electrification buildout (copper, NdFeB magnets) and the defense leg (REEs for missiles, radar, F-35, drones). The US response — government equity + price floors + offtake — is itself a circular state-funding structure, the mirror image of the private AI-capex loop, and it has spawned a wave of speculative junior-miner equities riding the strategic narrative.

1. The state-as-circular-financier structure (MP Materials)

The clearest case that the state plays the same circular role for minerals that NVIDIA/SoftBank play for OpenAI — only the recycling party is the US Treasury/DoD:

Mechanism: state takes equity + guarantees price + commits offtake → de-risks the supplier → supplier serves defense + commercial. Structurally a quasi-nationalization — the inverse of the private AI loop, same circularity.

2. The China chokepoint

China controls ~70% of mining and the dominant majority of refining/separation and magnet-making — and has demonstrated export controls as leverage (the 2023–25 gallium/germanium/antimony/REE restrictions). That is the binding constraint the Z3 defense_chokepoint proof formalizes (REE independence UNSAT until ~2028).

2b. Annual volumes — defense vs AI (added 2026-06-16, #48)

Quantifying the physical tonnage, and splitting demand between the defense leg (low-tonnage, high-criticality, no substitute) and the AI/electrification leg (high-tonnage bulk). Mine/processing shares are fact; the demand split is a labeled estimate.

MineralWorld mineChina shareThe chokepoint
Rare earths~390,000 t REO mined (2025, USGS o.o.m.)~70% mine, ~90% separation/magnets, ~100% heavy-REEthe binding constraint is separation, not ore — exactly the step weaponized Apr-2025
Antimony~83,000 t mined (2023, USGS)~48% mine, ~90% refinedthe US had no domestic mined antimony pre-Stibnite; Dec-2024 ban drove ~$25k→~$60k/t (~$51.8k Jun-2026)

3. The speculative tail

The strategic narrative has pulled a cohort of junior-miner equities (Critical Metals/Tanbreez, USA Rare Earth, Ucore, etc.) — real strategic need, but narrative-driven valuations; graded accordingly in the equities cohort (see macro-critical-minerals-equities).

4. Why on-thesis

Minerals are the physical floor under both legs the project maps: the AI buildout (magnets/copper for motors, data-center power) and the defense-industrial base. The state-circular financing is the macro-policy analogue of the private circular core — cross-ref geopolitics-defense-industrial-base, defense-rare-earth, macro-critical-minerals-equities.

Sources: CNBC — Pentagon largest MP shareholder; MP Materials 8-K (DoD price floor/offtake); Apple–MP $500M recycled-magnet release; CSIS/CFR/FDD on China export controls.

← Research index · structured data: macro-critical-minerals.json · macro-critical-minerals.md